A new group contribution-based model for estimation of lower flammability limit of pure compounds

In the present study, a new method is presented for estimation of lower flammability limit (LFL) of pure compounds. This method is based on a combination of a group contribution method and neural networks. The parameters of the model are the occurrences of a new collection of 105 functional groups. Basing on these 105 functional groups, a feed forward neural network is presented to estimate the LFL of pure compounds. The average absolute deviation error obtained over 1057 pure compounds is 4.62%. Therefore, the model is an accurate model and can be used to predict the LFL of a wide range of pure compounds.